基于改进布谷鸟算法的分数低阶盲均衡算法

A Fractional Lower Order Moment CMA Based on Improved Cuckoo Search

针对无线通信系统中传统常模盲均衡算法(CMA)在脉冲噪声环境下适应性较差,难以有效收敛的问题,提出了改进布谷鸟算法优化的分数低阶统计量常模盲均衡算法(SCS-FLOSCMA)。该算法将椋鸟鸟群的集体性行为引入到基础布谷鸟算法(CS)中,有效提高了搜索精度,减少了CS算法后期过早收敛的风险;然后把改进后的CS算法引入到分数低阶常模盲均衡算法(FLOSCMA)中,将搜索过后得到的全局最优巢作为均衡器的初始权向量。仿真表明,与CMA和FLOSCMA算法相比,该方法在均方误差曲线更稳定,收敛速度也更快。

In view of the large mean square error(MSE)and immerging in partial minimum easily for traditional constant modulus blind equalization algorithm(CMA)under impulse noise environment in wireless communication systems,a new blind equalization algorithm based on fractional lower order moment statistics and improved Cuckoo Search(SCS-FLOSCMA)was presented.By introducing the collective behavior of starling into Cuckoo Search(CS),the ability of global search was improved and the new algorithm could avoid local optimum.According to the characteristic of blind equalization algorithm.The improved CS algorithm found the optimum weight vector which fitness function was appropriate.Then the vector was used as the initial weight vector of constant modulus blind equalization algorithm based on fractional lower order moment statistics(FLOSCMA).Simulation results proved that,compared with the CMA and FLOSCMA,SCS-FLOSCMA had the faster convergence speed,the smaller MSE,and the clearer constellations of output signals.